Fig. 2

Illustration of the data processing pipeline and the issues detected during pre-processing. On panel (a) we show each step of data processing (white rectangles) with references to figures supporting the parameter selection at the actual processing step, and the raw (first blue rounded rectangle) and extracted (blue parallelograms) datasets shared along with this paper. Panels (b–e) demonstrate issues 2–4 and an example of the cleaned dataset. On these panels each black dot represents a signal, and color shaded areas show the problematic signals spotted by each procedure: (b) Issue 2 detected (red area) on a half-day file after initial data cleaning. The signals in the red area were subsequently removed; (c) Issue 3 detected (orange area) within the remaining signals from subplot b; (d) Issue 4 detected (yellow area) within the remaining signals from subplot c, and (e) remaining signals from subplot d constituting the final half-day file with cleaned signal sequence (i.e. pre-processed data).